once the initial diagnosis is made, all the subsequent evidence goes to elaborate on the initial diagnosis. The more one investigates the more it all fits together
You must have a point because I disagree with you on this subject so in a way that actually reinforces your point, at least as far as your own method of reaching conclusions goes.
I don't work that way. To me, I see both possibilities as equally valid until a thorough examination of all the pixels in the picture reveals either one image or the other as the prevalent one.
In the case of optical illusions, it is easy enough to see through them.
Take the one with the blocks and the wavy (or is it parallel) lines. At first you can't tell so you devise a test. Put a straight edge ruler against the screen of your PC monitor and the truth becomes obvious. They are parallel.
In the case of the pictures which can be one thing or another, the chances of each being correct remain equal. The conlusion that I reach is that the pictures are equally valid either way because they were either designed that way or some fluke of chance made them that way.
Conclusion = There is no correct way to look at them.